Dose response of a radiophotoluminescent glass dosimeter for TomoTherapy, CyberKnife, and flattening-filter-free linear accelerator output measurements in dosimetry audit.

PURPOSE: We experimentally determined the radiophotoluminescent glass dosimeter (RPLD) dose responses for TomoTherapy, CyberKnife, and flattening-filter-free (FFF) linear accelerator (linac) outputs for dosimetry audits in Japan. METHODS: A custom-made solid phantom with a narrow central-axis spacing of three RPLD elements was used for output measurement to minimise the dose-gradient effect of the non-flattening filter beams. For RPLD dose estimation, we used the ISO 22127 formalism. Additional unit-specific correction factors were introduced and determined via the measured data. For TomoTherapy (7 units) and CyberKnife (4 units), the doses were measured under machine-specific reference fields. For FFF linac (5 units), in addition to the reference condition, we obtained the field-size effects for the range from 5×5 cm to 25×25 cm. RESULTS: The correction factors were estimated as 1.008 and 0.999 for TomoTherapy and CyberKnife, respectively. For FFF linac, they ranged from 1.011 to 0.988 for 6 MV and from 1.011 to 0.997 for 10 MV as a function of the side length of the square field from 5 to 25 cm. The estimated uncertainties of the absorbed dose to water measured by RPLD for the units were 1.32%, 1.35%, and 1.30% for TomoTherapy, CyberKnife, and FFF linac, respectively. A summary of the dosimetry audits of these treatment units using the obtained correction factors is also presented. The average percentage differences between the measured and hospital-stated doses were <1% under all conditions. CONCLUSION: RPLD can be successfully used as a dosimetry audit tool for modern treatment units.

A Monte Carlo Study of Photon Beam Characteristics on Various Linear Accelerator Filters.

BACKGROUND: Intensity Modulated Radiation Therapy (IMRT) technique is an advanced method of radiotherapy leading into the development of Flattening Filter-Free (FFF) medical linear accelerators (Linacs). Monte Carlo simulation has been a standard method for calculation of particle transport due to precise geometry and material specifications. OBJECTIVE: This study is to obtain the design optimization of Flattening Filter Free (FFF) for 6 MV Linac machine. MATERIAL AND METHODS: In this simulating study, EGSnrc user code was used to simulate particles emitted from head of linac 6MV Varian to achieve the most suitable filter in FFF linac design. Monte Carlo simulation results of the PDD and profile, on the 10 × 10 cm2 field, were compared with the measurements. Differences in small profile beams from Monte Carlo simulation were also evaluated between FF and FFF linac. RESULTS: The spectrum on Monte Carlo simulation in isocenter was compared with Treatment Planning System (TPS) for each filter variation. The slight differences of average spectrum are simulated using 2 mm copper filter and FakeBeam with -1.52 ± 3.82% and -3.13 ± 3.61%. Whereas, for PDD and profiles, each variation has an average difference of 7.10 ± 0.70% and -5.99 ± 1.39%. CONCLUSION: FakeBeam filter is a proper filter for the use of linac design 6MV Varian. It is necessary to decrease the kinetic energy of electrons to perform MC simulations on FFF linac.

Interplay effect on a 6-MV flattening-filter-free linear accelerator with high dose rate and fast multi-leaf collimator motion treating breast and lung phantoms.

PURPOSE: Using a new linear accelerator with high dose rate (800 MU/min), fast MLC motions (5.0 cm/s), fast gantry rotation (15 s/rotation), and 1 cm wide MLCs, we aimed to quantify the effects of complexity, arc number, and fractionation on interplay for breast and lung treatments under target motion. METHODS: To study lung interplay, eight VMAT plans (1-6 arcs) and four-nine-field sliding-window IMRT plans varying in complexity were created. For the breast plans, four-four-field sliding-window IMRT plans were created. Using the Halcyon 1.0 linear accelerator, each plan was delivered five times each under sinusoidal breathing motion to a phantom with 20 implanted MOSFET detectors; MOSFET dose (cGy), delivery time, and MU/cGy values were recorded. Maximum and mean dose deviations were calculated from MOSFET data. The number of MOSFETs with at least 19 of 20 detectors agreeing with their expected dose within 5% per fraction was calculated across 106 iterations to model dose deviation as function of fraction number for all plan variants. To put interplay plans into clinical context, additional IMRT and VMAT plans were created and delivered for the sites of head and neck, prostate, whole brain, breast, pelvis, and lung. Average modulation and interplay effect were compared to those from conventional linear accelerators, as reported from previous studies. RESULTS: The mean beam modulation for plans created for the Halcyon 1.0 linear accelerator was 2.9 MU/cGy (two- to four-field IMRT breast plans), 6.2 MU/cGy (at least five-field IMRT), and 3.6 MU/cGy (four-arc VMAT). To achieve treatment plan objectives, Halcyon 1.0 VMAT plans require more arcs and modulation than VMAT on conventional linear accelerators. Maximum and mean dose deviations increased with increasing plan complexity under tumor motion for breast and lung treatments. Concerning VMAT plans under motion, maximum, and mean dose deviations were higher for one arc than for two arcs regardless of plan complexity. For plan variants with maximum dose deviations greater than 3.7%, dose deviation as a function of fraction number was protracted. CONCLUSION: For treatments on the Halcyon 1.0 linear accelerator, the convergence of dose deviation with fraction number happened more slowly than reported for conventional linear accelerators. However, if plan complexity is reduced for IMRT and if tumor motion is less than ~10-mm, interplay is greatly reduced. To minimize dose deviations across multiple fractions for dynamic targets, we recommend limiting treatment plan complexity and avoiding one-arc VMAT on the Halcyon 1.0 linear accelerator when interplay is a concern.

Dose evaluation of Grid Therapy using a 6 MV flattening filter-free (FFF) photon beam: A Monte Carlo study.

PURPOSE: Spatially fractionated radiotherapy is a strategy to overcome the main limitation of radiotherapy, i.e., the restrained normal tissue tolerances. A well-known example is Grid Therapy, which is currently performed at some hospitals using megavoltage photon beams delivered by Linacs. Grid Therapy has been successfully used in the management of bulky abdominal tumors with low toxicity. The aim of this work was to evaluate whether an improvement in therapeutic index in Grid Therapy can be obtained by implementing it in a flattening filter-free (FFF) Linac. The rationale behind is that the removal of the flattening filter shifts the beam energy spectrum towards lower energies and increase the photon fluence. Lower energies result in a reduction of lateral scattering and thus, to higher peak-to-valley dose ratios (PVDR) in normal tissues. In addition, the gain in fluence might allow using smaller beams leading a more efficient exploitation of dose-volume effects, and consequently, a better normal tissue sparing. METHODS: Monte Carlo simulations were used to evaluate realistic dose distributions considering a 6 MV FFF photon beam from a standard medical Linac and a cerrobend mechanical collimator in different configurations: grid sizes of 0.3 × 0.3 cm2 , 0.5 × 0.5 cm2 , and 1 × 1 cm2 and a corresponding center-to-center (ctc) distance of 0.6, 1, and 2 cm, respectively (total field size of 10 × 10 cm2 ). As figure of merit, peak doses in depth, PVDR, output factors (OF), and penumbra values were assessed. RESULTS: Dose at the entrance is slightly higher than in conventional Grid Therapy. However, it is compensated by the large PVDR obtained at the entrance, reaching a maximum of 35 for a grid size of 1 × 1 cm2 . Indeed, this grid size leads to very high PVDR values at all depths (≥ 10), which are much higher than in standard Grid Therapy. This may be beneficial for normal tissues but detrimental for tumor control, where a lower PVDR might be requested. In that case, higher valley doses in the tumor could be achieved by using an interlaced approach and/or adapting the ctc distance. The smallest grid size (0.3 × 0.3 cm2 ) leads to low PVDR at all depths, comparable to standard Grid Therapy. However, the use of very thin beams might increase the normal tissue tolerances with respect to the grid size commonly used (1 × 1 cm2 ). The gain in fluence provided by FFF implies that the important OF reduction (0.6) will not increase treatment time. Finally, the intermediate configuration (0.5 × 0.5 cm2 ) provides high PVDR in the first 5 cm, and comparable PVDR to previous Grid Therapy works at depth. Therefore, this configuration might allow increasing the normal tissue tolerances with respect to Grid Therapy thanks to the higher PVDR and thinner beams, while a similar tumor control could be expected. CONCLUSIONS: The implementation of Grid Therapy in an FFF photon beam from medical Linac might lead to an improvement of the therapeutic index. Among the cases evaluated, a grid size of 0.5 × 0.5 cm2 (1-cm-ctc) is the most advantageous configuration from the physics point of view. Radiobiological experiments are needed to fully explore this new avenue and to confirm our results.